A Capablanca Error

The book that I recommend more than any other is Chess Fundamentals by Jose Capablanca (1921).

Capablanca's book offers well-chosen examples with clear prose explanations. The structure of his work has influenced how I teach children. He starts with simple checkmates. The second section of chapter one presents pawn promotion and the idea of the opposition. Then, he offers slightly more involved pawn endings where the stronger side has two pawns to the weaker's one. He follows these basic checkmates and endings with middle game positions, then a discussion of the value of the pieces, then general opening principles.

Following chapter one, Capablanca repeats the sequence with more advanced lessons. Chapter two concerns endgames. Chapter three deals with planning in the middle game. Four develops middle game concepts that also have application to the opening and concludes with a model game. Five returns the endgame. The final chapter develops openings and middle games further. The book concludes with illustrative games drawn from his own play.

Capablanca's structure inspired my own rotation that I use teaching young players, whether through individual lessons or part of a club. I develop this structure in my ultimate camp workbook: Five Days to Better Chess (2017), available from Amazon. The five stage repeating process moves from checkmates and endings to middle games, openings, miniatures, and finally great games.

I have an old hardback edition of Chess Fundamentals published in 1934 that is in pretty good condition. I could protect it, but I use it. I also use a version that came free with one of those chess reader apps for the iPhone and iPad. I also have a database of all the positions from the book that someone put up on the web a couple of decades ago.

Yesterday morning, I was glancing through the database and decided to play Example 8 against the engine for practice.

White to move

After playing the position all the way to checkmate with Capablanca's method, I checked my solution with an engine. The computer found a faster checkmate.

Capablanca's solution begins with 1.Ke4. He claims that White cannot win with 1.f5 in view of 1...g6 and that the student should work this out.

The engine favors 1.f5 as the optimal solution.

All the years that I have been reading Capablanca's book, and I did not work out why 1.f5 fails. Once the computer informed that it was the winning line, however, I tried it against the engine and won easily. But, the engine did not attempt Capablanca's refutation, as that loses even faster.

As it turns out, the reasons 1...g6 does not refute 1.f5 are pretty easy to work out.

1.f5 g6 2.fxg6 Ke6 3.g5!

Perhaps this was the move overlooked by Capablanca. I'm not certain.

Black to move

3...Ke7 4.Ke5 Ke8 5.Ke6 Kf8 6.Kf6 Kg8

White to move

It is hard to believe that Capablanca would have believed this position to be a draw.

7.g7 Kh7

Maybe he saw this position in his head and not on a board, recognizing that 8.Kf7 would be stalemate.

8.g8Q+! Kxg8 9.Kg6

Black to move

This position is always a win no matter who has the move. It is also identical in important particulars to a position reached after the eighth move of Capablanca's sixth example (the pieces are on the e-file).

This error will not deter me from continuing to recommend Chess Fundamentals. Although not without small errors, it offers some of the best instruction available for those who have started to play chess are in need of fundamental principles of strategy and tactics.